Coexistence and Overlap: Comparing Azure Data Factory and Microsoft Fabric

Summary

The developer community is currently engaged in active discussions regarding the coexistence, differences, and functional overlap between Azure Data Factory (ADF) and Microsoft Fabric. While Microsoft Fabric represents the next generation of data integration as a unified SaaS platform, the classic ADF (PaaS) remains essential for many enterprises. Simultaneously, Microsoft is rolling out regular usability improvements to Fabric—such as the new column resizing feature in the SQL Analytics Endpoint—and expanding AI-driven workflows with Copilot capabilities.

What happened?

Recent articles and discussions across dev.to, LinkedIn, and tech portals have highlighted developer comparisons of features, pricing structures, and migration strategies between Azure Data Factory and Microsoft Fabric. Alongside these discussions, Microsoft introduced a practical usability update: developers can now manually resize column widths in the Microsoft Fabric SQL Analytics Endpoint results grid via drag-and-drop. Furthermore, updates regarding Microsoft Fabric Copilot and new Data Agents integrated with Microsoft 365 indicate a growing shift towards autonomous, agentic AI data workflows.

Why it matters

The transition from Azure Data Factory (PaaS) to Microsoft Fabric (SaaS) represents a fundamental shift in Microsoft’s data engineering approach. ADF requires managing infrastructure elements like Integration Runtimes and explicit publishing steps, but grants fine-grained network control. Conversely, Fabric simplifies operations by combining data integration, warehousing, and analytics into a single managed workspace. Small usability features like column resizing show that Microsoft is focused on reducing daily friction for developers using the platform.

Evidence

The growing momentum is supported by community posts and release notes:

Community Analysis: Tech blogs on dev.to detail the difference between ADF’s “Publish” model and Fabric’s “Save and Run” approach, alongside low-code alternatives.
Feature Release: A report on c-sharpcorner.com confirms the roll-out of Excel-like column resizing in the SQL Analytics Endpoint results grid, removing the need to export to Excel or SSMS just to view truncated text.
AI Advancements: Discussions on LinkedIn and official release summaries showcase how Fabric Copilot and Data Agents automate workflows across SQL and Eventhouse sources.

Analysis

The coexistence of ADF and Microsoft Fabric is expected to last for several years. While Fabric Data Factory simplifies development using Power Query-based Dataflow Gen2, it still lacks some of the highly customized CI/CD pipelines and complex orchestration capabilities of ADF. For existing Azure customers, this requires a balanced decision: Fabric is excellent for unified data gravity near Power BI and OneLake, while ADF remains the standard for isolated enterprise environments. However, the continuous injection of Copilot capabilities will make Fabric increasingly attractive over time.

Practical Takeaways

Start Evaluation: Teams should evaluate how well their current ETL workloads fit within Fabric’s SaaS architecture, especially if they are heavily integrated with Power BI.
Leverage New UI Grid: Developers should make use of the direct column resizing in the SQL Analytics Endpoint to inspect long strings, GUIDs, and JSON arrays without exporting.
Analyze Networking Needs: If your architecture depends heavily on private endpoints and custom virtual networks, maintain ADF as your primary orchestrator for the time being.

Open Questions

Will Microsoft achieve complete feature parity for all ADF pipeline activities in Fabric to allow automated, one-click migrations?
How will the actual costs of scaling Fabric capacities compare to the consumption-based pricing model of ADF for large-scale enterprise pipelines?